Group AffectSigal G. Barsade1 and Andrew P. Knight2

1Wharton School, University of Pennsylvania, Philadelphia, Pennsylvania 19104;email: barsade@wharton.upenn.edu2Olin Business School, Washington University in St. Louis, St. Louis, Missouri 63130;email: knightap@wustl.edu

Annu. Rev. Organ. Psychol. Organ. Behav. 2015.2:21–46

First published online as a Review in Advance onJanuary 2, 2015

The Annual Review of Organizational Psychologyand Organizational Behavior is online atorgpsych.annualreviews.org

This article’s doi:10.1146/annurev-orgpsych-032414-111316

Copyright © 2015 by Annual Reviews.All rights reserved

Keywords

emotion, teams, group dynamics, workplace

Abstract

Over two decades of research has indicated that group affect is an im-portant factor that shapes group processes and outcomes. We reviewand synthesize research on group affect, encompassing trait affect,moods, and emotions at a collective level in purposive teams. We be-gin by defining group affect and examining four major types of collec-tive affective constructs: (a) convergence in group affect; (b) affectivediversity, that is, divergence in group affect; (c) emotional culture; and(d) group affect as a dynamic process that changes over time. Wedescribe the nomological network of group affect, examining both itsgroup-level antecedents and group-level consequences. Antecedentsinclude group leadership, group member attributes, and interactionsbetween and relationships among group members. Consequences ofgroup affect include attitudes about the group and group-level co-operation and conflict, creativity, decision making, and performance.We close by discussing current research knowns, research needs, andwhat lies on the conceptual andmethodological frontiers of thisdomain.

21

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

INTRODUCTION

The body of research on group affect, which includes trait affect and collective moods andemotions, has developed significantly over the past 25 years. This development has led to an in-creased recognition not only that group affect exists, but that it is an essential piece of un-derstanding group dynamics (Barsade & Gibson 1998; Collins et al. 2013; George 1990, 1996;Kelly & Barsade 2001; Knight & Eisenkraft 2014). Group affect no longer lives on the fringes ofresearch on groups and teams; rather, it has become increasingly central to this domain. Scholarshave developed conceptual models and conducted empirical studies in the field and in the lab toexplore the antecedents, consequences, andmechanisms of group affect (Barsade&Gibson 2012,Collins et al. 2013). And yet, although there has been much progress, our understanding of thiscomplex phenomenon is still incomplete, and there remains significant room for additionalexploration.

We focus our review on theory and research involving affect, moods, and emotions at a col-lective level in purposive groups. By purposive group, wemean “an intact social system, completewith boundaries, interdependence for some shared purpose, and differentiated member roles”(Hackman & Katz 2010, p. 1210). Purposive groups are ubiquitous in contemporary organ-izations, charged with completing a broad range of tasks over diverse time horizons. A long-termmanufacturing team, a team developing a new piece of software over the course of a year, or anaction teamof healthcareworkers assembled to complete a surgical procedure in just a fewhours—all of these are purposive groups inwhich groupaffect canbe an intrinsic part of group functioning.Given the overlap between the literatures on work teams and purposive groups, we use the termsteam and group interchangeably throughout our review, which covers research conducted in boththe lab and the field.

We begin by first addressing the question, What is group affect? We then highlight fourprominent types of affective constructs at the collective level. Next, we review how theorists andresearchers have mapped the nomological network of group affect—that is, the web of rela-tionships between group affect and other group-level antecedents and consequences.We concludeby summarizing a fewparticularly consistent findings of research on group affect, identifying areasin which additional research is most needed, and highlighting promising ideas on the frontiers oftheory and research on collective affect.

WHAT IS GROUP AFFECT?

In their conceptualization of group affect and review of the literature, Barsade & Gibson (1998)characterized group affect in two basic ways. The first was what they termed a “top-down”approach, in which group affect was viewed as “a whole,”with characteristics and properties ofthe group acting upon the emotions of the individuals within it. The second process, which theytermed a “bottom-up” approach, manifested as group affect that emerged from the “sum of itsparts” and was the result of the aggregate of individual group members’ affective states andtraits. Kelly & Barsade (2001) further developed this conceptualization, integrating severalconstructs into a comprehensive model of group affect that delineated the order and processesthroughwhich bottom-up and top-down forces work in concert to create emotion at a collectivelevel. Importantly, theirmodel identified both implicit and explicit affective transfer processes—including emotional contagion, behavioral entrainment, and vicarious affect—that can serve asconduits for transferring affect among group members. They also discussed how the group’scontext, such as technology, physical space, and intergroup relations, could also influence groupaffect and vice versa.

22 Barsade � Knight

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

Combining these approaches, we review theory and research on a number of constructs andprocesses that are most relevant for scholars interested in studying the affect of a group. In muchthe sameway that the term affect is an umbrella term that encompasses trait affect, state affect, anddiscrete emotion (Ashforth & Humphrey 1995, Barsade & Gibson 2007, Young 1961),throughout our reviewweuse the termgroupaffect broadly, as an umbrella term that encompassesseveral different forms of collective-level affective constructs.

GROUP AFFECT: COLLECTIVE-LEVEL AFFECTIVE CONSTRUCTS

Existing research has examined a range of affective constructs. We highlight the most prominentforms of group affect and review empirical studies that have examined these conceptualizations.Specifically, from the bottom-up, compositional perspective, group affect can manifest as (a)convergence and (b) affective diversity, that is, divergence in group members’ individual affect.From the top-down perspective, group affect can manifest as (c) emotional culture and (d) groupaffect as a dynamic process that changes over the course of the group’s life span. Table 1 sum-marizes these different affective constructs.

Affective Group Constructs as Convergence in Individual Group Member Affect

The most well-studied form of group affect, by far, is as an affective experience that is shared, orheld in common, by themembers of a group or team. In this conceptualization, a collective positiveor negativemood emerges in a group because each groupmember feels a similar level of individualpositive or negative mood; that is, individual group members converge in their affective experi-ences at a given point in time. A conceptualization of group affect as a shared construct gainedtraction in organizational behaviorwith thework ofGeorge (1990, 1995, 1996). George’s (1990,1996) discussion of group affect, or group affective tone, specified convergence in individualfeelings as a necessary precondition for conceptualizing collective group affect.

George (1990) emphasized the role of compositional effects to explain how andwhy individualgroup members’ affective states converge. Specifically, George drew from Schneider’s (1987)attraction-selection-attrition model to argue that (a) people seek out work groups composed ofsimilar others with respect to affect, (b) organizations and work groups choose to bring aboardpeople who are similar in affect, and (c) those members of a work group most dissimilar in affectare most likely to turn over. Because of these selection effects, group composition becomes in-creasingly homogeneous over time with respect to affect (George 1996). George’s (1990) findingsregarding the relative homogeneity of state affect inwork groupswere consistentwith the idea thatlong-standing groups are characterized by a unique, homogeneous collective affect.

The attraction-selection-attritionmodel is particularly useful in explainingwhy themembers oflong-term groups converge in affect. However, the majority of theory and research to date onhomogeneous affect has focused on convergence in mood (i.e., state affective experience) ratherthan trait affect. Specifically, research on mood convergence has focused on the mechanisms in-volved in momentary affective transfer processes (e.g., Elfenbein 2014, Kelly & Barsade 2001),such as emotional contagion and similar group member reactions to shared events (Weiss &Cropanzano 1996). Drawing from theory and research on primitive emotional contagion(i.e., Hatfield et al. 1993, 1994), which involves the largely automatic and subconscious transfer ofemotions from person to person, organizational researchers have examined the mechanisms thatunderlie affective transfer across a broad spectrum of groups working in a wide range of contexts(e.g., Barsade 2002, Bartel & Saavedra 2000, Totterdell et al. 1998). Emotional contagion occursthrough automatic mimicry of the facial expressions, voice, and body movements of others (e.g.,

23www.annualreviews.org � Group Affect

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

Dimberg et al. 2000, Hess & Fischer 2014, Lundqvist &Dimberg 1995). This mimicry then leadsthe perceiver to actually feel the emotion, effectively catching the emotion of the other person.

Initial research on affective convergence in work groups focused on documenting the existenceof the phenomenon and exploring some of the antecedents and consequences of mood at a col-lective level. For example, using an experience sampling approach, Totterdell et al. (1998) foundemotional convergence in field settings, with a significant, positive relationship between individualmoodandgroupmoodover time innursing andaccounting teams, controlling for sharedproblemsthat group members faced together. In an experimental study of emotional contagion amongbusiness student groups engaging in a groupmanagerial negotiation, Barsade (2002) manipulatedthe mood expressed by a trained confederate, tracking emotional contagion in the group withminute-by-minute video-coder ratings, showing how one group member’s mood can infect themoods of other group members and lead to a shared group mood. Several additional studiessupport the basic idea that individuals tend to converge in their affective experiences during groupinteractions. For example, in a field study of 70 work teams, Bartel & Saavedra (2000) docu-mented convergence in group members’moods. And more recent research has extended this ideaby showing how other variables, such as individual differences or attitudes toward the group, maystrengthen or weaken the tendency for affect to transfer among group members (e.g., Ilies et al.2007), a topic we discuss in greater detail below.

Affective Group Constructs as Affective Diversity: Divergence in Individual GroupMember Affect

Convergence in individual group member affect has been the most widely studied conceptuali-zation of an affective construct at the collective level; however, it is not the only manifestation ofgroup affect. Although George’s (1990, 1996) conceptualization emphasized consistency andsimilarity in affective experiences among work group members, other early conceptual models ofthe emergence of group affect noted that, even though groups do have a natural tendency toconverge, a range of forces acts upon them, which can lead groupmembers to feel differently fromone another at any given point in time (e.g., Barsade & Gibson 1998, Kelly & Barsade 2001).Relative to theory and research on affective convergence in groups, however, there has been lessresearch on affective diversity, or divergence, in groups. This is surprising, given the amount ofattention that organizational scholars have paid to diversity in other individual attributes (vanKnippenberg & Schippers 2007, Williams & O’Reilly 1998).

Table 1 Group affect: collective-level affective constructs

Collective-level affective

constructs Description

Affective convergence Affect that is shared, or held in common, by the members of a group or team

Affective diversity/divergence

The configuration of affect across group members that is not shared, nor held in common, bymembers of a group or team

Emotional (affective)culture

Behavioral norms, artifacts, and underlying values and assumptions reflecting the actual expressionor suppression of the discrete emotions comprising the culture and the degree of perceivedappropriateness of these emotions, transmitted through feeling and normative mechanisms withina group

Affective dynamism The pattern of change over time in group affect owing to processes that lead collective affect to emergeand fluctuate

24 Barsade � Knight

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

Barsade et al.’s (2000) field study of affective diversity in top management teams of large andprominent organizations (i.e., Fortune 500 companies) stands as an exception to the heavy focuson convergence in the group affect literature. Also grounding their predictions in similarity-attraction theory (i.e., Byrne 1971), the authors proposed competing hypotheses about theinfluence of affective diversity on group functioning. They examined the consequences of di-versity in team members’ trait positive affectivity and found that positive affective diversity wassignificantly related to team processes (i.e., related positively to top management team conflictand negatively to team cooperativeness) and marginally negatively related to the top man-agement team outcome of corporate financial performance. Barsade et al.’s analysis furtherindicated that the impact of diversity on team processes and outcomes was moderated by theaverage trait positive affectivity of the team (although, as predicted, negative trait affectivity hadno influence on these outcomes). In a study of nuclear power plant crews performing a crisissimulation, Kaplan et al. (2013) found effects similar to those reported by Barsade et al. (2000).Specifically, Kaplan et al. found that diversity in trait positive affectivity hindered group ef-fectiveness. A mediating mechanism that connected affective diversity to performance was thedegree to which participants experienced negative emotion during the crisis simulation, withgreater diversity in trait positive affectivity leading to greater negative emotion and, ultimately,to poorer performance.

Recent lab-based research has opened additional promising directions for understanding themeaning and effects of affective diversity or divergence in affectwithin groups. In a series of studiesthat manipulated the degree of affective diversity in pictures of groupmembers,Magee & Tiedens(2006) found that external observers judged groups inwhich therewas greater diversity in emotionamong the groupmembers as sharing less of a common fate and holding less shared responsibilityfor group outcomes. Their study suggested that the degree of affective diversity among people is anindicator that people use when judging the “groupiness” of a group.Weisbuch&Ambady (2008)took a different approach to studying affective diversity, focusing on the explanatory process ofemotional contagion. They proposed and found that when focusing on an out-groupmember, theprocess of contagion can lead to affective divergence, rather than affective convergence. In a similarvein, VanKleef and colleagues’ (VanKleef 2009,VanKleef et al. 2010)model of emotions as socialinformation highlights the idea that contextual factors, such as whether individuals encounter oneanother in a cooperative or a competitive context, shape whether they will converge to a commonaffective experience or diverge in their feeling states. Last, Elfenbein’s (2014) model of affectivelinkage highlights, among other effects, the nuanced interplay of affect and group categorizationthat can lead to affective convergence or divergence.

In contrast to models grounded in similarity-attraction theory and divergent emotional con-tagion, which suggest that affective diversity will impede group functioning, a few theoreticalarticles have proposed that divergence in group member affect could enhance group performanceon tasks that require divergent thinkingand creativity. Tiedens et al. (2004) proposed a conceptualmodel of the antecedents and consequences of group affective diversity, which they referred to as“emotional variation.” Adopting a top-down and bottom-up perspective of group affect that issimilar to Barsade and colleagues’ models of group affect (i.e., Barsade & Gibson 1998, Kelly &Barsade 2001), Tiedens et al. highlighted that compositional and contextual forces can yielddiversity in group members’ affect in certain situations related, for example, to social hierarchiesand differentiated role structures.

More generally, scholars are just beginning to explore divergent group affect, a topic essentialto understanding behavior in teams. Notwithstanding theoretical arguments regarding the ben-efits of diversity in affect (e.g., George & King 2007, Tiedens et al. 2004), the little empiricalresearch that scholars have published to date (i.e., Barsade et al. 2000, Magee & Tiedens 2006)

25www.annualreviews.org � Group Affect

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

indicates that differences in group member affect likely disrupt the functioning of purposivegroups, adversely impacting group effectiveness.

Emotional (Affective) Culture

A key top-down force that drives group affect is emotional culture (based on normative groupforces; also called affective culture) (Barsade&Gibson 1998, 2007, 2012; Kelly&Barsade 2001).The most comprehensive definition of the construct is the “behavioral norms, artifacts andunderlying values and assumptions reflecting the actual expression or suppression of (the discreteemotions comprising the culture1) and the degree of perceived appropriateness of these emotions,transmitted through feeling and normative mechanisms within a social unit” (Barsade & O’Neill2014, p. 558).

Drawing from the literature on affect management, Barsade & Gibson (1998) and Kelly &Barsade (2001) highlighted the role of normative processes in groups that are inherent to theemotional culture definition. The study of these normative processes began with Hochschild’s(1983) classic study of the emotional labor of flight attendants, and the idea that emotionalculture is the set of shared norms, sometimes prescribed by management, that governs howorganizational members express affect. A related concept of emotional display norms or rules(Ekman 1973) in groups was explored in a variety of contexts, including bank tellers andsalespeople (Rafaeli & Sutton 1989), bill collectors (Sutton 1991), Disney employees (VanMaanen & Kunda 1989), McDonald’s counter employees (Leidner 1993), nurses (Diefendorffet al. 2011), and physicians (Smith & Kleinman 1989). Emotional display norms constitute onemanifestation of emotional culture, and the intersection of an individual’s emotional experi-ences and expressions with emotion norms prescribed by the group resonated strongly in thefield (Ashforth & Humphrey 1995).

A recent longitudinal field study at a large long-term care facility was one of the first toexplicitly examine the construct of emotional culture as more formally and broadly defined(Barsade&O’Neill 2014). In this study, the authors predicted and found that a stronger cultureof companionate love—the behavioral norms, artifacts, and underlying values and assumptionsreflecting the actual expression or suppression of affection, caring, compassion, and tenderness—predicted greater employee teamwork and satisfaction and reduced employee emotional ex-haustion and absenteeism. Furthermore, they found that this culture of love among employeesrippled out to influence patients and their families as well. In units with a stronger culture ofcompanionate love, patients experienced more positive moods, reported a higher quality of life,had more satisfied family members, and required fewer trips to the emergency room. Anotherrecent study examined the impact of emotional culture on both work and nonwork outcomes infire stations, finding that an emotional culture of joviality and an emotional culture of com-panionate love complemented one another, relating to less employee risk-taking behavior outsideof work (O’Neill & Rothbard 2014). An emotional culture of joviality also had an independenteffect on teamperformance andwas associatedwith increased coordination andalso increased risktaking on the job.

1Barsade&O’Neill (2014) examined and defined the emotional culture of companionate love, andwe have extrapolated fromtheir definition to emotional culture more broadly.

26 Barsade � Knight

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

Dynamic Views: The Ebb and Flow of Affect in Groups over Time

Although the majority of empirical research has focused on static group affect, Kelly &Barsade (2001) emphasized dynamic affective emergence—how, over time, the nature ofcollective affect can change. They integrated the effects of affective (emotional) culture andcomposition, suggesting that group affect, over time, is dynamic and a product of thesebottom-up and top-down forces. Specifically, they posited that momentary affective expe-riences feed back into the group’s history, shaping the appraisals that group members make offuture events and experiences. In recent years, theorists have further expanded upon the ideathat momentary experiences of group affect become inputs into future group affectiveexperiences (Hareli & Rafaeli 2008, Walter & Bruch 2008), and some, although not much,empirical research has examined the dynamic interplay of affect and group memberexperiences.

Walter & Bruch (2008), for example, proposed a cyclical conceptual model of group positiveaffect in which similarity of positive affect (i.e., the extent to which group members converge inpositive affect) is reciprocally related to the quality of interpersonal relationships among groupmembers. High-quality relationships, Walter & Bruch (2008) posited, facilitate emotional con-tagion and convergence in group positive affect, and convergence in group positive affect furtherenhances the quality of interpersonal relationships among group members. Through these re-ciprocal relationships, group positive affect and group relationship quality are embedded ina virtuous cycle. In a related conceptualization of cyclical group affect, Hareli & Rafaeli (2008)argued that individualmoods and emotions influence themembers of a group, sparking emotionalreactions and responses thatmay be consistentwith either convergence (i.e., through contagion) ordivergence (i.e., through reactivity) in mood. These reactions to an initial emotional expressionthen further perpetuate and spiral through the group, contributing to an ebb and flow of affectiveresponses across group members over time.

Last, although there has been very limited empirical research on affective dynamics in groupsand teams, several authors have suggested that affect may play a role in how groups and teamschange over time (e.g., Barsade & Gibson 2007, George 2002, Gersick 1991, Knight 2015,Moreland 1987, Tuckman 1965). Knight (2015), for example, argued that group affect influenceshow group members regulate their activity over time when facing a deadline. In a field study ofmilitary teams facing a deadline in a prestigious physical and tactical competition, Knight foundthat shared positive mood at the temporal midpoint of the project prompted a decline in teammembers’ focus on an exploration of new strategies, thus facilitating an effective developmentaltrajectory and enhanced team performance. Shared negative mood at the midpoint, in contrast,sustained team members’ exploratory efforts even as their deadline drew near and thus hinderedteam performance.

The dearth of empirical research on patterns of affective dynamics over time in teams is a sig-nificant limitation of existing theory and research on group affect. Existing conceptualizations ofgroup affect all highlight that affect and group processes are dynamically intertwined as groupsmove through time (e.g., De Dreu et al. 2001, George 2002, Kelly & Barsade 2001). Furthermore,theorists have speculated that feedback loops play a prominent role in understanding the role thatgroup affect plays in group functioning (e.g., Hareli & Rafaeli 2008, Walter & Bruch 2008).Without empirical research on group affect over time, however, the temporal dynamics of groupaffect and group processes remain relatively unknown. More research is needed to explore howa group’s affective history is built over time and how momentary affective experiences in groupsshape subsequent affective experiences.

27www.annualreviews.org � Group Affect

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

Summary

Above we review four different constructs of affect at the collective level, two of which focus ongroup affect stemming from aspects of group composition and two of which focus on group affectas a contextual factor that influences group members (and summarize these four constructs inTable 1). The convergence of individual dispositions or moods has been the most heavily studiedmanifestation of group affect to date. Compelling theoretical models and some empirical evidencesuggest, however, that other group-level affective constructs meaningfully influence group pro-cesses and outcomes. In particular, affective diversity should receive further empirical attention inthe future. From a top-down perspective, a greater focus on the affective values, norms, and deepunderlying assumptions of differing types of emotional culture will add an important perspectiveto the group affect (and culture) literatures. Last, the least developed aspect of group affect,understanding the dynamic interplay of group affect andother group characteristics, inwhich eachinfluences the other, is absolutely critical to gaining a complete understanding of the nature ofaffect in groups and teams.

THE NOMOLOGICAL NETWORK OF AFFECTIVE GROUP CONSTRUCTS:ANTECEDENTS

As described above, one primary focus of early theory and research on affect in groups wasexplaining and documenting how and why affect emerges at a collective level. We review herethree antecedents of the emergence of collective-level affective constructs that have garnered themost attention from researchers to date: (a) group leadership, (b) attributes and attitudes of groupmembers, and (c) relationships and interactions among group members.

Group Leadership

Several studies suggest that formal leaders have a disproportionate impact on the nature of affectthat emerges in a group or team. By occupying a position of high visibility in the group and holdingformal power, which may dictate patterns of group interactions, group leaders are likely to infectthe members of a group with their individual affective states. Furthermore, through their roles,leadersmay shape the group’s emotional culture with respect to if, when, and how groupmembersshould express their experienced moods and emotions during group interactions. For thesereasons, leadership theorists and researchers suggest that management of the group’s affectivedynamics is a key function that group leaders fulfill (e.g., Humphrey 2002, Pescosolido 2002).

In support of the idea that leaders influence the development of group affect, George (1995)found in an early field study of sales managers that leader positive mood was positively andsignificantly related to the collective positive mood of the sales group. Furthermore, both leaderpositive mood and the positive mood of the group were positively associated with sales groupperformance. Relatedly, Sy et al. (2005) used an experimental design in which they manipulatedgroup leader mood to investigate the influence of leader affect on group affect, processes, andoutcomes. Consistent with George’s (1995) findings and prior research on emotional contagion(i.e., Barsade 2002), Sy et al. (2005) found that leader mood significantly influenced the shared,collective mood of the group and also was related to group processes and outcomes. Other studieshave replicated these results, showing that leaders infect the members of their groups with theiraffective states, driving the nature of group affect that emerges (e.g., Chi et al. 2011, Johnson 2009,Seong&Choi 2014). To organize and extend these findings, Sy&Choi (2013) recently proposedthe leader activation and member propagation model, in which leaders spark processes of mood

28 Barsade � Knight

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

contagion, which then ripple out and propagate among the members of the group. Sy & Choi’smodel accounts for compositional aspects of the group (e.g., group diversity, the similarity of theleader and group members) and personal attributes of team members (e.g., susceptibility toemotional contagion) to explain the nuances of when and how shared group mood emerges.

One notable area in which group leadership and affect seem to be particularly tightly inter-twined is research on charismatic leadership. Recent studies of leader charisma have found thataffect is a key mechanism through which charismatic leaders influence group performance (e.g.,Bono & Ilies 2006, Erez et al. 2008, Pastor et al. 2007, Sy et al. 2013). Erez et al. (2008), forexample, argued that the characteristics of charismatic leaders lead them to both experience andexpress vivid, positive, high-arousal emotions, which are easily transferred to followers and leadfollowers to experience shared positive moods. Specifically focused on groups, Sy et al. (2013)proposed and found reciprocal relationships among leader charisma and group positive andnegative mood; specifically, emotionally expressive leaders cultivated shared group mood andwere perceived as charismatic. Additionally, group positivemood enhanced immediate judgmentsabout leader effectiveness, which contributed to subsequent perceptions of leader charisma.Negative mood largely had the opposite effects, decreasing followers’ perceptions of leader ef-fectiveness and charisma. Together, these studies suggest that moods and emotions may be themedium through which charisma influences the attitudes, motivation, and behaviors of groupmembers.

Last, there have been some interesting preliminary investigations of moderators of the re-lationship between a leader’s mood and the characteristics of a team. For example, Van Kleef et al.(2009) found that leader anger led to better team performance for teams with lower average levelsof agreeableness, whereas leader happiness led to better team performance for teams with higheraverage levels of agreeableness. As Sy&Choi’s (2013) model similarly suggested, the relationshipbetween the leader’s affective displays and group composition is one promising direction forresearch on the intersection of group affect and leadership.

Personality Traits and Demographic Characteristics of Individual Group Members

In addition to documenting a tendency for the members of groups to converge in affect duringgroup interactions, researchers have also shown that the extent to which there is group conver-gence or divergence depends in part on individual group member personality traits and de-mographic characteristics. With respect to personality characteristics, researchers have drawnfrom Doherty’s (1997) conceptualization of individual differences in susceptibility to emotionalcontagion to argue that individuals highly susceptible to contagion are more likely to share af-fective experiences with their teammates than are individuals who are less susceptible (i.e., Ilieset al. 2007, Sy & Choi 2013). Ilies et al. (2007) also proposed that collectivistic tendencies—a relatively stable individual difference—moderate the link between individual and group affect. Ina field study of teams, these authors found support for their predictions that greater collectivistictendencies would lead tomore contagion. In an experimental study, Barsade (1995) explored howadditional personality characteristics (specifically, self-monitoring) influenced the extent to whichgroupmembers converged toward the mood of the group. She found that high self-monitors weremore likely to become infected with the mood displayed by a confederate than were low self-monitors. Most recently, Sy & Choi (2013) found that group diversity in extraversion andneuroticism, personality characteristics that have affective components within them (Watson2000), inhibits convergence in group member affective states.

Although several studies have examined how personality differences shape the emergence ofgroup affect, research on the intersection of individual demographic attributes and the emergence

29www.annualreviews.org � Group Affect

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

of group affect is more limited. Totterdell and colleagues (i.e., Totterdell 2000, Totterdell et al.1998) showed that the demographic attributes of group members moderate the link betweenindividual affect and the collective affect that emerges in a group. Specifically, Totterdell andcolleagues’ studies found that group member age is related to the linkage between the individualand the group, with older groupmembersmore likely to be prone to emotional contagion from thegroup. Taking a different approach, Hentschel and colleagues (2013) examined perceptions ofteam diversity as an antecedent to group affect, proposing and finding that when group memberssee their teamashighly diverse, they experiencemore sharednegative feelings. These findings are inlinewithmore general social-functional perspectives on affect in groups (e.g., Fischer&Manstead2008, Keltner&Haidt 1999), which highlight that sharing affective states may have enabled earlyhumans tomore readily identify in-group from out-groupmembers. If individuals differ on salientdemographic attributes, a resulting categorization of one another as out-group members mightinhibit convergence in their affective states. Some research on the mimicry of emotional statesprovides support for these ideas (e.g., van der Schalk et al. 2011, Weisbuch & Ambady 2008).However, the findings of Magee & Tiedens (2006) suggest that similarity in affect can overrideprominent demographic characteristics (i.e., gender, race) in shaping individuals’ perceptions ofgroup boundaries. Further research is needed to tease apart the links between demographicdifferences, personality differences, and collective affect.

Aside from individual demographic and personality characteristics, research also suggests thatindividual attitudes toward the group—specifically, the extent to which a group member feelspositively about or is committed to his or her group—maymoderate the link between an individualgroup member’s affect and the collective-level affect of the group. Totterdell et al. (1998) foundthat teammemberswhoweremore committed to their team and perceived their team environmentmore positively had greater affective convergence. In a study of cricket teams during a competition,Totterdell (2000) replicated the finding that group members who are highly committed to thegroup are more likely to share affective experiences with other group members. In addition,Tanghe et al. (2010) found in two studies, one survey based and another scenario based, that groupidentification is positively related to the convergence of group member affect.

Existing research thus suggests that there is meaningful variation in the extent to which groupmembers are influenced by the affective experiences of others in the group. Individual charac-teristics—including personality attributes, demographic attributes, and individual attitudes to-ward the group—moderate the relationship between an individual groupmember’s affect and thecollective-level affect of the group.

Relationship Structure and Frequency of Interactions Among Group Members

Foundational theories of emotional contagion (e.g., Hatfield et al. 1994) describe how severalmechanisms, including facial mimicry, emotional comparison, and empathy, contribute toconvergence in affect in groups. Extending these ideas, researchers have found that variant typesand structural patterns of interactions between and relationships among groupmembers influencethe extent to which group members converge in affect at a given point in time.

In their study of affective convergence in work groups, Bartel & Saavedra (2000) found thatconvergence in mood among group members is associated with both task and social in-terdependence: the more interdependent the group, the more convergence in group members’moods. Furthermore, these authors found that membership stability and mood regulation normspredicted mood convergence. In his study of cricket teams, Totterdell (2000) also examined theunderlying assumption of interdependence among group members, which is key to most explicitmechanisms of emotional contagion. He found that during times of interdependent, collective

30 Barsade � Knight

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

action throughout the course of the cricket match, convergence in team members’ affect wasgreatest. In experimental work that provides further support for these ideas, Klep et al. (2011,2013) showed that manipulations of affect in which group members interacted interdependentlyyielded stronger effects on group processes than did manipulations of affect that individual groupmembers experienced in isolation.

Reflecting the importance of interdependence for emotional contagion, Totterdell et al. (2004)theorized that interpersonal relationships among group members act as conduits through whichaffect flows in groups. These authors found support for the importance of social network ties foraffective convergence in work groups in data from workers in a vehicle manufacturing company.In particular, employees’ work ties and their structural equivalence were positively related tosimilarity of mood. Furthermore, the overall structure of the network, in terms of size and density,was related to the moods that workers experienced. Moving beyond face-to-face interactionpatterns, Cheshin et al. (2011) examined the emergence of shared group mood among groupmembers working together virtually in a computer-based negotiation task and found that bothtext-based and behavior-based cues lead to emotional contagion (even in the absence of direct, in-person interactions).

Thus, there is support for within-group variation in how tightly connected an individual groupmember’s affect is to the collective group’s affect based on the patterns of interactions within thegroup and the structure of relationships among groupmembers. Themore interconnected a groupmember is with others in the group, the more likely it is that he or she will share affectiveexperiences with others.

THE NOMOLOGICAL NETWORK OF AFFECTIVE GROUP CONSTRUCTS:CONSEQUENCES

There is a growing literature on the consequences of group affective constructs. To date, however,there are some consistent and some contested findings regarding how specific forms of group affectrelate to group processes and outcomes (Collins et al. 2013, Kozlowski & Ilgen 2006). Below wereview research on how group affective constructs influence four broad categories of groupprocesses and outcomes: (a) attitudes, cognitions, and behavior toward the group; (b) memberinteractions, cooperation, and conflict; (c) group creativity and decision making; and (d) groupeffectiveness and performance.

Group Member Attitudes, Cognitions, and Behavior Toward the Group

Given the close conceptual and empirical associations between affect and satisfaction at the in-dividual level (i.e., Brief & Weiss 2002, Locke 1976), it is not surprising that group researchershave similarly examined the relationship between group affect and outcomes such as groupmorale, satisfaction, and commitment. Much like the findings at the individual level, researchersstudying groups have found a positive relationship between positively valenced affective con-structs and thepositive attitudes thatmembers hold toward their groups. For example, in a studyofleadership and teams, Chi et al. (2011) found that team members who shared positive affectiveexperiences together (i.e., teams high in team positive mood) were more satisfied with their teams,more committed to their teams, and engaged in a greater degree of helping behavior thanmembersof teamswho did not share common positive affective experiences. As described above, Barsade&O’Neill (2014) found that a stronger emotional culture of companionate love among staffmembers in long-term care units predicted greater employee satisfaction and teamwork. Similarlyreflecting the behavioral manifestations of these attitudes, a few studies (George 1990, Mason &

31www.annualreviews.org � Group Affect

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

Griffin 2003) have shown that shared positive feelings among group members decrease with-drawal behaviors, such as absenteeism (Barsade & O’Neill 2014). Somewhat differently, Gibson(2003, p. 2153) examined the relationship between shared group positive affect and the formationof group efficacy, “a group’s collective belief in its capacity to perform a task.” In a lab-basedsimulation and a field study,Gibson (2003) found support for the idea that shared positive affect ingroups,which promotes optimismandactivates positive cognitions about group experiences in thepast, increases group efficacy.

Relative to studies of collective positive affective constructs, there has been less evidence of theinfluence of negative collective affective constructs on group member attitudes, cognitions, andbehavior toward the group. However, in one study of negative discrete group emotion, Duffy &Shaw (2000) found that mean group envy—that is, the degree to which each group member feltenvy toward other members of the group—had a negative influence on group cohesiveness andgroup potency, which in turn were positively related to group member satisfaction andperformance.

Member Interactions: Cooperation, Conflict, and Coordination

Group affect has been found to influence group dynamics, including group cooperation, conflict,and coordination. Drawing from theory and research on the influence of positive affect on in-terpersonal relations, Barsade et al. (2000) found that the top management team’s trait positiveaffect and the diversity in the team’s trait positive affect interacted to shape intrateam interactions,including team conflict and cooperativeness. The form of the interaction indicated that highlydiverse teams low in average trait positive affect experienced the most significant adverse out-comes. In an experimental study of emotional contagion in groups, Barsade (2002) also found thatpositive emotional contagion was positively related to multiple measures—self-report ratings,video-coder ratings, and team member cooperative allocation of funds—of intragroup co-operativeness and negatively related to intragroup conflict in a leaderless group negotiation task.Negative emotional contagion had the opposite result. Providing further support for the asso-ciation of group affect and group conflict, Choi & Cho (2011) proposed and found that groupnegative affect serves a critical bridging role between task conflict and relationship conflict inworkgroups.

Grawitch et al. (2003b) used a mood-induction approach in a laboratory study to examine thedifferential impact on group processes and outcomes of shared positive, shared negative, andneutral moods. The results of this study suggested that the members of groups in a positive moodcondition were more involved in the task and engaged in a greater degree of consensus-seekingbehavior, relative to the members of groups in a negative or neutral mood condition. Relatedly, inthe lab-based study mentioned above, Sy et al. (2005) found that groups led by an individual forwhom a positive mood had been induced showed higher shared positive group mood and werebetter coordinated, compared to groups led by an individual for whom a negative mood had beeninduced. Interestingly, the authors found that groups experiencing shared negative feelingsexpended more effort on the task than did groups experiencing shared positive feelings.

In general, the positive findings reported above are consistent with Kelly & Barsade’s (2001)model of group affect, which posited that the way in which group members interact with oneanother and their affective experiences are reciprocally related. In groups in whichmembers sharepositive affective experiences, group interaction patterns are likely to be positive and cooperative.Negative affect, however, has led to more varied results. Although there are some effects ofnegative emotional contagion on group dynamics (e.g., Barsade 2002, Sy et al. 2005), positiveaffect generally has a more direct influence on group processes and outcomes when examining

32 Barsade � Knight

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

group convergence or divergence. This comports with the evidence that there is a minimal sig-nificant relationship between trait negative affectivity and interpersonal outcomes, and traitnegative affectivity instead influences more intrapersonal variables, such as stress (e.g., Barsadeet al. 2000, Grawitch et al. 2003b, McIntyre et al. 1991, Watson & Pennebaker 1989, Watsonet al. 1992). Indeed, McIntyre et al.’s (1991, p. 67) review indicated that trait positive affect, butnot trait negative affect, is related to “diverse indicators of social activity and interpersonalsatisfaction,” whereas trait negative affect, but not trait positive affect, is related to “somaticcomplaints, psychopathology, and self-reported stress.”

Knight & Eisenkraft’s (2014) meta-analysis of research on group positive and negative affectprovides additional explanation for mixed findings regarding the effects of group affect on in-terpersonal dynamics and performance. Grounded in a social-functional perspective, Knight &Eisenkraft (2014) found that positive affect has consistent positive effects on group social in-tegration and task performance; in general, groups that share positive feelings are cohesive andperform at a high level. In contrast, the effects of negative affect on social integration and taskperformance are sensitive to contextual factors, including the source of affect (external or internalto the group) and the life span of the group (one-shot or ongoing). Shared negative feelings thatstem from external sources or occur in one-shot groups promote social integration and taskperformance, whereas shared negative feelings that stem from internal sources or occur in ongoinggroups undermine social integration and task performance. The results of this meta-analysisindicate that group positive affect has broad and diffuse benefits for social integration in groups,whereas the effects of negative affect are more sensitive to situational contingencies.

Group Creativity and Decision Making

Mirroring the significant interest in how affective experiences influence cognition and creativity inindividuals (e.g., Damasio 1994, George 2007, Isen & Baron 1991, Lazarus 1982, Zajonc 1980),group theorists and researchers have examined how different forms and types of collective affectdrive group creativity and decision making. And, mirroring the disagreements in the individual-level literature on affective influences on cognition, creativity, and decision making, group-levelresearch has mostly revolved around the question of whether positively or negatively valencedgroup affect enhances group performance on tasks requiring group decision making or creativity.

The theorizing behind group creativity and decision-making research has been grounded inindividual-level theoretical arguments about affect and creativity. That is, positive affect promotesenhanced cognitive flexibility in choosing the best way to problem solve, innovate, and makedecisions to fit the problem at hand (Isen 2000). It also increases variety-seeking behavior andbroadens cognition (e.g., Fredrickson 1998). Starting with creativity, in an experimental study ofgroups engaged in a brainstorming task, Grawitch et al. (2003a) found that group positive mooddirectly positively influenced the originality of group ideas, relative to neutral group mood. Theyalso found that groups in which members were induced to share a positive affective experienceoutperformed groups in either a negative or neutralmood condition on a creativity task.However,there are occasional studies showing that negative group mood can positively influence creativity(Jones &Kelly 2009). For example, in a survey-based study of 68 Chinese R&D teams, Tsai et al.(2012) argued that both shared team positive mood and shared team negativemood impact groupcreativity and found a complex set of relationships between positive and negativemood and groupcreativity, in which team trust served as a boundary condition. Focusing on a different dimensionof group affect, Knight & Baer (2014) found that group arousal promoted higher creative per-formance in brainstorming groups by increasing the degree to which group members built uponand extended one another’s ideas.

33www.annualreviews.org � Group Affect

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

With regard to decisionmaking, the results are evenmore complex—with much of the researchconducted within the paradigm of the distributed information task (i.e., group members possessunique information) that requires group members to combine their uniquely held information tosolve a murder mystery. Using this type of task, Bramesfeld & Gasper (2008) found across twoexperiments that group positive mood was positively related to decision-making quality. How-ever, using the same type of task, Kooij-de Bode et al. (2010) found that groups induced to feelnegatively performed better on a decision-making task when information was distributed acrossgroup members, compared to groups in a positive mood condition. They determined that in-formation elaboration explains the influence of group negative mood on decision-making quality.Similarly, vanKnippenberg et al. (2010) found that groups induced to experience a negativemoodin this distributed information task engaged in greater information elaboration, relative to groupsinduced to experience a positive mood, which led to enhanced decisionmaking. There was also aninteraction between the mean trait negative affect of the group and the mood of the group,predicting information elaboration.The authors interpreted the interactionas evidence for the ideathat trait affect can override the effects of state affect in groups. Some studies using a non-distributed information paradigm have found an influence of group mood on decision makingaround the allocations of group resources (Barsade 2002), but to more fully understand thephenomenon of affect and group decisionmaking as awhole, researchers need to adopt additionaland more varied stimuli and types of decisions.

Research thus currently paints an ambiguous picture of how group affect influences groupdecision making and creativity. One stream of research, grounded in individual-level theory andresearch showing themore established beneficial effects of people engaged in decision-making andcreativity tasks (Lyubomirsky et al. 2005), suggests that positive group affective experiences yieldenhanced group performance on such tasks. A second stream of research, grounded in individual-level theory and research suggesting that negative affect can promote persistence and criticalthinking, indicates that negative group affective experiences can promote the sharing of uniqueinformation among group members, leading to better outcomes. It is likely that the influence ofgroupaffect ondecisionmaking and creativity is not straightforward anddirect; rather, howgroupaffect influences decision making and creativity likely depends on contextual factors (George2011). As evidence of how context can influence the effects of group affect, Knight (2015) foundthat the relationship between group affect and strategic exploration in teams changed over thecourse of time as teams approached a deadline. Positive affect early in a team’s life span promotedexploration, with team members seeking alternative ways of completing team tasks. As teamsapproached the deadline, however, positive affect inhibited exploratory search as teams turnedtheir focus to implementation. Additional research, both in the field and in the lab, is needed totease apart the nuanced effects of group affect on group decision making and creativity. This isparticularly so given the importance of groups as information processors in organizations (Hinszet al. 1997); further investigation of the influence of positive and negative affect on collectivecognition in groups is sorely needed.

Group Effectiveness and Performance

In addition to research on how collective-level affective constructs influence group effectiveness ondecision-making and creativity tasks, group researchers have also proposed and found that groupaffect influences how well groups perform on other types of tasks. For example, George (1990)gathered survey data from 26 work groups and found that the mean trait negative affect of thegroup was positively related to the shared negative groupmood that groupmembers experienced.This shared group negative mood was related to less prosocial behavior toward customers and

34 Barsade � Knight

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

greater employee absenteeism (with some evidence of shared grouppositivemoodbeing negativelyrelated to absenteeism). In an examination of how leader and group affect influence sales per-formance in a similar retail setting,George (1995) also found a significant relationship between theshared group positive mood of sales groups and better sales performance. On the sports field,among professional cricket players, Totterdell (2000) found that team-level positive moodpositively related to individual-level performance asmediated by the players’ individual moods. Inthe lab, positive collective mood resulting from emotional contagion led to greater ratings ofeffective performance on the part of the other members in the group, with the opposite results fornegative collective mood (Barsade 2002). Sy et al. (2005) found that leaders who transmittedpositive mood also had groups with more successful coordination (but less expended effort), asmediated by group positive mood (Sy et al. 2005).

With respect to negative affect, in a study of 61 work teams in a multinational automotivecomponentmanufacturer,Cole et al. (2008) found that mean negative affective tone (measured byamore generalmeasure of asking employees about their negative affect atwork) directly negativelyinfluenced teamperformance, as rated by team supervisors. Teamnegative affective tonemediatedthe relationship between dysfunctional teambehavior and teamperformance, and the relationshipbetween negative affective tone and performancewasmoderated by nonverbal expressivity.Whenteam members showed what they were feeling, the relationship was stronger between negativeaffective tone and performance than when they did not show their feelings.

At the organizational level, in a study of leadership, team affect, and team performance,Hmieleski et al. (2012) proposed that shared authentic leadership indirectly influences teamperformance by eliciting shared positive emotions among team members. Using survey datagathered from the leaders of small businesses, the authors found a positive, significant relationshipbetween team positive affect, defined as converged positive affective experiences among teammembers, and the performance of new ventures (i.e., revenue, employment growth). Barsade et al.(2000) also found an influence on organizational outcomes, with greater trait positive affectivediversity among senior management teams in major organizations marginally related to lesssuccessful firm financial performance.

Several scholars have suggested that group affect indirectly influences group performancethrough some of the mechanisms described above, such as intragroup interactions and groupattitudes and beliefs. For example, Chi et al. (2011) found that team positive mood indirectlyinfluenced teamperformance through teamgoal commitment, team satisfaction, and teamhelpingbehaviors. Knight (2015) found that shared team positive mood at the temporal midpoint of theteam’s life span indirectly influenced teamperformance by shaping how teams paced their focus oncoming up with alternative approaches to their tasks. Knight also reported that the mean traitpositive affect of the teamwas significantly positively related to objective teamperformance aboveand beyond a range of team characteristics, including team ability, team experience, and earlyplanning activities. In theirmeta-analysis of the effects of group affect, Knight&Eisenkraft (2014)found that social integration partially mediates the effects of group positive affect and groupnegative affect on group task performance, with group positive affect promoting task performancethrough social integration and group negative affect having nuanced, moderated effects on taskperformance through social integration. Last, in a laboratory study of brainstorming groups,Knight&Baer (2014) found that information elaborationmediated the effects of group arousal ongroup performance.

In a study of the impact of a specific, positive discrete emotion shared by team members onperformance, Rhee (2006) experimentally examined the role of shared group joy in team func-tioning. Drawing from Fredrickson’s (1998) broaden-and-buildmodel of positive emotions, Rheepredicted and found that teams induced to experience joy were more effective, in part because of

35www.annualreviews.org � Group Affect

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

interactions characterized by broaden-and-build behaviors. Additionally, Barsade & O’Neill(2014) found that the emotional culture of companionate love among staff members predicteda critical set of performancemetrics for the organization—companionate lovewas associatedwithincreased patient positive mood, improved patient quality of life, fewer unnecessary emergencyroom visits for residents, and enhanced satisfaction of patients’ families.

In summary, existing research on the relationship between group affective constructs andgroup effectiveness suggests that group affect—and, specifically, shared group positive affect—isdirectly and indirectly related to general group effectiveness. The effects of group negative affect,however, are more ambiguous and seem to be dependent on situational contingencies (Knight &Eisenkraft 2014). In a narrow range of situations, group negative affect might enhance groupperformance; however, in general, group negative affect seems to hinder group task performance.Significantly more research is needed, however, to understand how different forms and types ofcollective-level affect influence group effectiveness across a range of tasks and group contexts.

KNOWNS AND NEEDS IN THEORY AND RESEARCH ON GROUP AFFECT

Although much has been learned, there is still much unknown. Below we highlight those pre-dictions or assertions for which our knowledge is relatively clear—what we call the researchknowns. Then we underscore several key predictions, assertions, or types of empirical researchthat are most needed to further advance theory and research on group affect—what we call theresearch needs. The sidebar summarizes these research knowns and research needs.

Research Knowns

Over the past 25 years of research, one key finding has emerged consistently—affect in groupsdevelops toward homogeneity. When a group of people work together with one another, researchsuggests that it is likely that individual group members will converge in their affective states,leading to shared collective-level affect. This tendency has been supported by research both in thefield (e.g., Bartel&Saavedra2000,George1990,Totterdell 2000,Totterdell et al. 1998) and in thelab (Barsade 2002, Sy & Choi 2013). These studies have relied on self-reports of affect (e.g.,Barsade 2002, George 1990, Knight 2015, Totterdell et al. 1998), as well as observer ratings ofaffect (e.g., Barsade 2002, Barsade &O’Neill 2014, Bartel & Saavedra 2000), and have includedboth newly formed groups (e.g., Barsade 2002, Sy et al. 2005) and groupswith a long prior history(e.g., Bartel & Saavedra 2000, George 1990, Totterdell et al. 1998).

As reviewed by a number of authors, there are several forces that pull group members towardhomogeneity in affective experiences (Barsade & Gibson 1998, 2012; George 1996; Kelly &Barsade 2001; Parkinson et al. 2005). First, because mood is interpersonally contagious, teammembers may infect one another with their mood (Barsade 2002, Totterdell 2000, Totterdell et al.1998), contributing to a convergence in mood. Second, owing to attraction-selection-attritionprocesses in long-standing groups or departments, the members of a team may have similar traitaffective dispositions, leading them to react and interact in similar ways (Barsade et al. 2000,George 1996, Kelly & Barsade 2001). Third, team members may feel similarly at a given point intime because, during the course of their work, they encounter similar external stimuli (Weiss &Cropanzano 1996) or are governed by the same emotional culture regarding affective expression(Barsade & Gibson 1998, 2007; Barsade & O’Neill 2014; Kelly & Barsade 2001). In summary,convergence in affectmay be akin to affective homeostasis in groups. In the short term, exposure tocommon events and the forces of emotional contagion pull groupmembers’ affective states toward

36 Barsade � Knight

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

one another. In the long term, the forces of attraction-selection-attrition in addition to behavioralnorms further unify group affect.

Another research known emerging from empirical research on group affect is that sharedpositively valenced group affect facilitates the development of positive group attitudes and pro-cesses, such as commitment, satisfaction, and viability (e.g., Barsade & O’Neill 2014, Chi et al.2011, Grawitch et al. 2003b), as well as cooperative group behavior and social integration (e.g.,Barsade 2002, Barsade et al. 2000, Grawitch et al. 2003b, Knight & Eisenkraft 2014). Sharednegatively valenced group affect, alternatively, seems to have no simple, main effect (Grawitchet al. 2003b, McIntyre et al. 1991, Watson et al. 1992) or, when there is a main effect, seems toimpede these same group attitudes and processes (e.g., Barsade 2002,George 1990). The influenceof negative affect on these group attitudes and processes seems to be more sensitive to situationalcontingencies (Knight&Eisenkraft 2014). Furthermore, with the exception of decision-making orcreativity tasks, for which extant research is more ambiguous regarding the influence of groupaffect, positively valenced group affect seems to be consistently positively associated with groupperformance (e.g., Barsade 2002, Barsade & O’Neill 2014, Barsade et al. 2000, Chi et al. 2011,Knight 2015, Knight & Eisenkraft 2014, Rhee 2006). The broad conceptual model underlyingthese findings is generally consistentwith the input-process-output framework that has dominatedthe literature on groups and teams (Hackman & Katz 2010, Kozlowski & Ilgen 2006). Positivemoods and emotions, theorists and researchers have suggested, indirectly influence group andteam outputs through their relationships with positive attitudes and coordinated, cooperativegroup behavior.

Additionally, current research suggests that there is a tight link between group leadership andthe nature and valence of group affect that emerges (Sy & Choi 2013). Displayed group leadermoods and emotions have been shown to influence the experiencedmoods and emotions of groupmembers (e.g., George 1995, Hmieleski et al. 2012, Sy&Choi 2013, Sy et al. 2005), consequentlyinfluencing group attitudes, processes, and outcomes. Especially of note, charismatic leadershipseems tightly linked to affective processes in teams, so much so that scholars have suggested thataffect is a primary conduit through which charismatic leaders influence their followers and theirgroups (e.g., Bono & Ilies 2006, Erez et al. 2008, Pastor et al. 2007, Sy et al. 2013). Despite theseconsistent findings, quantitative evidence nevertheless is needed to substantiate and help explainwhy leaders have a disproportionate impact on the affect that emerges in groups as compared toother, nonleader, members of the group.

Research Needs

Below we highlight three research needs that, if addressed, would significantly advance scholars’understanding of affect in groups. First, more research is needed about the causes and con-sequences of affective diversity in group members’ dispositional (trait) affect, moods, andemotions. Despite growing theoretical attention to affective diversity in groups and teams, em-pirical evidence to validate, challenge, and extend conceptual models remains more limited. Thereis some empirical evidence indicating that affective diversity in groupmembers’ affective traits andstates can interfere with positive group functioning (e.g., Barsade et al. 2000, Kaplan et al. 2013).This findingmay relate to the new line of research examining divergence in emotional contagion(e.g., Elfenbein 2014, Hess & Fischer 2014, Moody et al. 2007, van der Schalk et al. 2011,Weisbuch&Ambady 2008). Yet some authors have suggested in theoretical models that affectivediversity may benefit groups engaged in decision-making tasks requiring complex thinking,minority influence, and creativity (i.e., George&King 2007, Tiedens et al. 2004). Reasoning thatdivergence in group member mood states might provoke or stimulate divergence in cognition,

37www.annualreviews.org � Group Affect

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

these authors suggested that convergence in affect may lead groups to reach premature con-clusions.However, empirical research is needed to explore these predictions andbetter understandhowaffective diversity influences processes andoutcomes in groups engaged in a variety of tasks. Itis possible that, similar to other forms of diversity (i.e., Milliken & Martins 1996), affectivediversity is a double-edged sword—the benefits of divergent thinking may come at the cost ofgroup social integration.

Second, real-time, process-oriented research is needed on the ebb and flow of affect, moods,and emotionswithin groups and teams over time. The individual-level empirical research on affectis beginning to shift toward more dynamic research, using, in particular, experience samplingtechniques to capture how individuals’ affective states change over time and how changing af-fective states influence individual and workplace outcomes (e.g., Alliger & Williams 1993,Amabile et al. 2005, Ilies & Judge 2005, Rothbard &Wilk 2011). This process-focused researchapproach has enabled scholars to test predictions about how affective dynamics evolve over timeand also to examine the causal relationships that link affect to workplace outcomes, such asmotivation and creativity. Although these types of reciprocal models based on feedback loops are

GROUP AFFECT: RESEARCH KNOWNS AND NEEDS

Research Knowns

n Affect in groups develops toward homogeneity. When a group of people work together, it is likely thatindividual group members will converge in their affective states, leading to shared collective-level affect. Thiscan be due to

n Emotional contagion,n Attraction-similarity-attrition, orn Shared affective experiences.

n With regard toworkplace outcomes, such as commitment, satisfaction, and viability, aswell as cooperative groupbehavior and social integration, creativity, decision making, and performance,

n Shared, positively valenced group affect generally facilitates the positive development of all of the above.n Shared negatively valenced group affect generally facilitates the negative development of all of the above;however, it is sensitive to situational contingencies that can then in combination lead to positive outcomes onthe processes listed above.

n There is a tight link between a group leader’s affect and the affect of group members.

Research Needs

n More research is needed about the causes and consequences of affective diversity—divergence in groupmembers’dispositional (trait) affect, moods, and emotions.

n Real-time, process-oriented research is needed on the ebb and flow of affect, moods, and emotions within groupsand teams over time. Though difficult to do, a variety of methodological techniques (e.g., experience sampling inthe field, video coding, and computer applications) can be helpful with this.

n Amoremulticultural orientation is needed, with greater investigation of howoverarching societal factors, such asculture, can influence the nature and influence of affect within and between groups.

38 Barsade � Knight

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

often discussed theoretically (e.g., Hareli &Rafaeli 2008, Kelly& Barsade 2001,Walter& Bruch2008), there is little process-oriented, longitudinal research to validate core theoretical predictionsregarding group affect. Although it is methodologically challenging, understanding how group-level affect changes over time, reciprocally influencing other group-level constructs, is critical foradvancing theory and research on affect in groups.

Third, research is needed to understand how overarching societal factors, such as societalculture, might influence the nature and effects of group affect. Although group affect researchhas been conductedwith samples from a variety of nations (e.g., the United States, South Korea),and researchers have used samples from different societies to test the generalizability of findings(e.g., Gibson 2003), there are, to our knowledge, no direct cross-cultural examinations thatexplore how the nature and effects of group affect might vary across cultures. Ilies et al. (2007)examined how individualism and collectivism shape individuals’ tendencies to convergetoward the mood of their teammates; however, the authors examined these factors as in-dividual differences. Given that groups are always embedded within a cultural context—andthat the effects of group affect are likely context dependent (e.g., Elfenbein 2007)—the dearthof cross-cultural theory and research on group affect is a significant limitation of the existingliterature. Theoretical hypotheses and empirical examinations about these cultural contextsare much needed to ensure that this literature is relevant to organizations operating across theglobe.

PRACTICAL IMPLICATIONS

The research knowns articulated above suggest a few recommendations for managers involvedin creating and leading groups and teams. First, existing research has shown that group com-position, specifically composition with respect to group member trait affectivity, is a significantdriver of the shared feelings that emerge in a group over time. This suggests that, when creatinggroups and teams, managers should be attentive to group members’ affective dispositions.Second, existing research has shown that the affective states that group leaders express sig-nificantly shape the experienced affective states of groupmembers. Thus, in addition to shapinggroup affect through personnel selection, group leaders also can shape group affect and, in-directly, group functioning through their own emotional expressions. Several studies havedemonstrated the power of subtle manipulations of leader expressions of affect in altering groupdynamics (e.g., Sy et al. 2005, Van Kleef et al. 2009). A practical implication that emerges fromexisting research is thus that leaders can use their own affective expressions to influence thedynamics of their groups.

But what kind of group affect should leaders and managers try to cultivate in groups andteams? For many types of tasks, all else equal, cultivating shared positive feelings among groupmembers is likely to benefit group functioning and group performance. Although existing researchis less clear regarding the effects of group affect on groupperformance for decision-making tasks, itis noteworthy that Knight & Eisenkraft’s (2014) meta-analysis of group affect research foundoverall positive effects of group positive affect on group performance, across tasks. Thus, the stateof existing research suggests that promoting positive affective states in groupsmaybe beneficial forgroupdynamics andperformance. This is not to say that shared negative feelings are not functionalfor groups. Indeed, for a group that has just lost out on a big contract to a competitor, anger couldbe a useful short-term energizing emotion for the team. However, the current literature suggeststhat shared negative feelings promote healthy group functioning in a narrower range of situationsthan do shared positive feelings.

39www.annualreviews.org � Group Affect

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

ON THE FRONTIERS

At the frontiers of the group affect literature, there are exciting conceptual and methodologicaldevelopments. First, an emerging area is the conceptualization of affective skills, also known asemotional intelligence, as collective, group constructs. Although research to date is limited, a fewconceptual and empirical articles suggest that groups themselves can vary with respect to emo-tional monitoring, regulation, and other affective competencies or skills. Druskat &Wolff (2001,p. 133) offered an early conceptualization of group emotional intelligence, which they defined as“the ability to develop a set of norms that manage emotional processes so as to cultivate trust,group identity, and group efficacy.”As evidenced by their definition, they viewed group emotionalintelligence as instrumental for enabling groupmembers to develop and sustain collective beliefs—specifically, trust, identity, and efficacy—that theory and research suggest are critical drivers ofgroup effectiveness. Elfenbein (2006) andCôté (2007) further elaboratedon the construct of groupemotional intelligence, exploring both the nature of the construct and its antecedents and con-sequences. Elfenbein (2006), for example, argued that group emotional intelligence can emerge indifferent compositional forms, such as through convergence or diversity in group members’ in-dividual emotional intelligence. Different forms of group emotional intelligence, Elfenbein sug-gested, influence group processes and outcomes in different ways.

Group-level conceptualizations of affective skills and competencies are currently empiricallylimited to group emotional intelligence. The field also needs to examine other types of skills relatedto collective-level affective competencies, such as team emotion recognition accuracy (Elfenbeinet al. 2007) or emotional aperture, the ability of individuals to read collective emotions (Sanchez-Burks & Huy 2009).

Second, technological and statistical developments present group researchers with novel andeffective tools to examine affective dynamics in groups and teams over time. In the growing field ofaffectivecomputing—an area of theory, research, and practice dedicated to enhancing the capacityfor computers to read and express emotions (Picard 2000)—scholars from a range of disciplines,including computer science and engineering, machine learning, biology, and psychology, arecollaborating to design and implement novel methods for measuring individuals’ moods andemotions and for analyzing how thesemoods and emotions change over time. For example, Picardand colleagues (i.e., Picard et al. 2001, Poh et al. 2010) have developed wireless, unobtrusivesensors thatmeasure activation of the sympathetic nervous system,which group researchersmightuse to understand how groupmembers’ activation influences group processes and outcomes (e.g.,Knight & Baer 2014). There are systems that code filmed facial expressions (D’Arcey 2013) inways similar to Ekman & Friesen’s (2003) well-known FACS rubric (although these systemscurrently focus only on one face at a time). Researchers have also developed software solutions tocode text-based communications, such as instant messages, for affective constructs (e.g., Bollenet al. 2011). Because these measures leverage technology, they often enable real-time and con-tinuous recording of affective states, which might facilitate examinations of affective dynamics atwork. And, because they are often less obtrusive than existing measurement approaches, theseemerging technologies might be helpful in examining the role of implicit affect that is “activated orprocessed outside of conscious awareness that [influences] ongoing thought, behavior andconscious emotional experience” (Barsade et al. 2009, p. 139). As Barsade et al. (2009) reviewed,implicit affect is a growing area in psychology in understanding the basis of people’s cognition,motivation, and behavior. Although it is a nascent field within organizational behavior, sub-conscious affective group processes could have powerful implications for the group-level con-sequences, such as performance, that we discuss above.

40 Barsade � Knight

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

Last, because we focus our review on collective-level affective constructs in purposive groups,we have not incorporated into this article the stream of theory and research on intergroup emotion(e.g., Mackie et al. 2000), which examines the antecedents and consequences of individuallyexperienced emotions that are activated as a function of group-based identities (Niedenthal &Brauer 2012). However, one could posit that the same psychological processes that occur whengroup-based identities are activated intrapsychically also occur, and perhaps more strongly (e.g.,Shteynberg et al. 2014a,b),when individuals are actually facedwith others in their groupwhohavediffering group-based identities. As such, it could be useful to have more cross integration amongthese two research streams.

CONCLUSION

As the affective revolution in organizational behavior continues (Barsade et al. 2003), theory andresearch on collective-level affect in groups and teams are gaining attention. As shown in thisreview, despite extensive theoretical conceptualizations of group affect, empirical studies havetaken a relatively more narrow course, with empirical research focusing predominantly ona conceptualization of group affect as the homogeneity of affective experiences across groupmembers. Although important, we see a singular focus on homogeneity in group members’ affectas limiting to our understanding of the collective affective experience that occurs in groups. Weencourage researchers to rigorously test the diverse conceptualizations of collective constructs—such as affective diversity, affective dynamics, and emotional culture—that scholars have pro-posed, so as to deepen our understanding of this important phenomenon within organizationalbehavior.

DISCLOSURE STATEMENT

The authors are not aware of any affiliations, memberships, funding, or financial holdings thatmight be perceived as affecting the objectivity of this review.

ACKNOWLEDGMENTS

Many thanks go to Inhyo Choi, Rebecca Silver, Kevin Sweeny, and Kimberly Yao for their re-search assistance. The authors would also very much like to thank the collaborators with whomthey have conducted joint research in the quest to better understand the influence of group affect inorganizational life: Markus Baer, Noah Eisenkraft, Donald E. Gibson, Janice R. Kelly, JochenMenges, Olivia A. O’Neill, Jaime M. Potter, Jeffrey Sonnenfeld, Jean D.F. Turner, Danielle V.Tussing, and Andrew J. Ward.

LITERATURE CITED

Alliger GM, Williams KJ. 1993. Using signal-contingent experience sampling methodology to study work inthe field: a discussion and illustration examining task perceptions andmood. Person. Psychol. 46:525–49

Amabile TM,Barsade SG,Mueller JS, StawBM.2005.Affect and creativity atwork.Adm. Sci.Q. 50:367–403Ashforth BE, Humphrey RA. 1995. Emotion in the workplace: a reappraisal. Hum. Relat. 48:97–125Barsade SG. 1995. Emotional contagion in groups. Diss. Abstr. Int. A Humanit. Soc. Sci. 56(5-A):1876Barsade SG. 2002. The ripple effect: emotional contagion and its influence on group behavior.Adm. Sci. Q.

47:644–75

41www.annualreviews.org � Group Affect

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

Barsade SG, Brief AP, Spataro SE. 2003. The affective revolution in organizational behavior: the emergence ofaparadigm. InOrganizationalBehavior: TheState of the Science, ed. JGreenberg, pp. 3–52.Mahwah,NJ:Erlbaum Assoc. 2nd ed.

Barsade SG,GibsonDE. 1998.Group emotion: a view from top andbottom. InResearchonManagingGroupsand Teams, Vol. 1, ed. MA Neale, EA Mannix, pp. 81–102. Greenwich, CT: JAI

Barsade SG, Gibson DE. 2007. Why does affect matter in organizations? Acad. Manag. Perspect. 21:36–59Barsade SG, Gibson DE. 2012. Group affect: its influence on individual and group outcomes. Curr. Dir.

Psychol. Sci. 21:119–23Barsade SG, O’Neill OA. 2014. What’s love got to do with it? A longitudinal study of the culture of com-

panionate love and employee and client outcomes in the long-term care setting. Adm. Sci. Q. 59:551–98Barsade SG, Ramarajan L, Westen D. 2009. Implicit affect in organizations. Res. Organ. Behav. 29:135–62Barsade SG,WardAJ, Turner JDF, Sonnenfeld JA. 2000. To your heart’s content: amodel of affective diversity

in top management teams. Adm. Sci. Q. 45:802–36Bartel C, Saavedra R. 2000. The collective construction of work group moods. Adm. Sci. Q. 45:197–231Bollen J, Mao H, Zeng X. 2011. Twitter mood predicts the stock market. J. Comput. Sci. 2:1–8Bono JE, Ilies R. 2006. Charisma, positive emotions, and mood contagion. Leadersh. Q. 17:317–34Bramesfeld KD, Gasper K. 2008. Happily putting the pieces together: a test of two explanations for the effects

of mood on group-level information processing. Br. J. Soc. Psychol. 47:285–309BriefAP,WeissHM.2002.Organizational behavior: affect in theworkplace.Annu.Rev. Psychol.53:279–307Byrne DE. 1971. The Attraction Paradigm. New York: AcademicCheshin A, Rafaeli A, Bos N. 2011. Anger and happiness in virtual teams: emotional influences of text

and behavior on others’ affect in the absence of non-verbal cues. Organ. Behav. Hum. Decis. Process.116:2–16

Chi N, Chung Y, Tsai T. 2011. How do happy leaders enhance team success? The mediating roles oftransformational leadership, group affective tone, and team processes. J. Appl. Soc. Psychol. 41:1421–54

Choi K, Cho B. 2011. Competing hypotheses analyses of the associations between group task conflict andgroup relationship conflict. J. Organ. Behav. 32:1106–26

Cole M, Walter F, Bruch H. 2008. Affective mechanisms linking dysfunctional behavior to performance inwork teams: a moderated mediation study. J. Appl. Psychol. 93(5):945–58

Collins AL, Lawrence SA, Troth AC, Jordan PJ. 2013. Group affective tone: a review and future researchdirections. J. Organ. Behav. 34:S43–62

Côté S. 2007. Group emotional intelligence and group performance. In Affect and Groups, ed. EA Mannix,MA Neale, CP Anderson, pp. 309–36. Res. Manag. Groups Teams Vol. 10. Greenwich, CT: JAI

Damasio AR. 1994.Descartes’ Error: Emotion, Reason, and the Human Brain. New York: Grosset/PutnamD’Arcey JT. 2013.Assessing the validity of FaceReader using facial EMG. PhDThesis,Calif. StateUniv.,ChicoDe Dreu CKW, West M, Fischer A, MacCurtain S. 2001. Origins and consequences of emotions in or-

ganizational teams. In Emotions at Work: Theory, Research, and Applications for Management, ed.RL Payne, CL Cooper, pp. 199–217. New York: Wiley

Diefendorff JM, Erickson RJ, Grandey AA, Dahling JJ. 2011. Emotional display rules as work unit norms:a multilevel analysis of emotional labor among nurses. J. Occup. Health Psychol. 16(2):170–86. doi:10.1037/a0021725

Dimberg U, Thunberg M, Elmehed K. 2000. Unconscious facial reactions to emotional facial expressions.Psychol. Sci. 11:86–89

Doherty RW. 1997. The Emotional Contagion Scale: a measure of individual differences. J. Nonverbal Behav.21:131–54

Druskat VU, Wolff SB. 2001. Group emotional intelligence and its influence on group effectiveness. InThe Emotionally Intelligent Workplace, ed. C Cherniss, D Goleman, pp. 132–55. San Francisco:Jossey-Bass

DuffyMK, Shaw JD. 2000. The Salieri syndrome: consequences of envy in groups. Small Group Res. 31:3–23Ekman P. 1973. Darwin and Facial Expression: A Century of Research in Review. New York: AcademicEkmanP, FriesenWV. 2003.Unmasking the Face: AGuide toRecognizingEmotions fromFacial Expressions.

Los Altos, CA: Malor Books

42 Barsade � Knight

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

Elfenbein HA. 2006. Team emotional intelligence: what it can mean and how it can affect performance. InLinking Emotional Intelligence and Performance at Work: Current Research Evidence with Individualsand Groups, ed. V Druskat, F Sala, G Mount, pp. 165–84. Mahwah, NJ: Erlbaum Assoc.

Elfenbein HA. 2007. Emotions in organizations: a review and theoretical integration. Acad. Manag. Ann.1:315–86

Elfenbein HA. 2014. The many faces of emotional contagion: an affective process theory of affective link.Organ. Psychol. Rev. 4:326–62

Elfenbein HA, Polzer JT, Ambady N. 2007. Team emotion recognition accuracy and team performance.In Research on Emotions in Organizations, Vol. 3: Functionality, Intentionality and Morality, ed.NM Ashkanasy, WJ Zerbe, CEJ Härtel, pp. 87–119. Greenwich, CT: JAI

Erez A, Misangyi VF, Johnson DE, LePine MA, Halverson KC. 2008. Stirring the hearts of followers:charismatic leadership as the transfer of affect. J. Appl. Psychol. 93:602–15

Fischer AH, Manstead ASR. 2008. Social functions of emotion. InHandbook of Emotions, ed. M Lewis,JM Haviland-Jones, LF Barrett, pp. 456–68. New York: Guilford. 3rd ed.

Fredrickson BL. 1998. What good are positive emotions? Rev. Gen. Psychol. 2:300–19Gibson CB. 2003. The efficacy advantage: factors related to the formation of group efficacy. J. Appl. Soc.

Psychol. 33:2153–86George JM. 1990. Personality, affect, and behavior in groups. J. Appl. Psychol. 75:107–16George JM. 1995. Leader positive mood and group performance: the case of customer service. J. Appl. Soc.

Psychol. 25:778–94George JM. 1996. Group affective tone. In Handbook of Work Group Psychology, ed. M West, pp. 77–93.

New York: WileyGeorge JM. 2002. Affect regulation in groups and teams. InEmotions andWork, ed. R Lord, R Klimoski,

R Kanfer, pp. 183–217. San Francisco: Jossey-BassGeorge JM. 2007. Creativity in organizations. Acad. Manag. Ann. 1:439–77George JM. 2011. Dual tuning: a minimum condition for understanding affect in organizations. Organ.

Psychol. Rev.1(2):147–64George JM, King EB. 2007. Potential pitfalls of affect convergence in teams: functions and dysfunctions of

group affective tone. In Affect and Groups, ed. EA Mannix, MA Neale, CP Anderson, pp. 97–123. Res.Manag. Groups Teams Vol. 10. Greenwich, CT: JAI

Gersick CJ. 1991. Revolutionary change theories: a multilevel exploration of the punctuated equilibriumparadigm. Acad. Manag. Rev. 16:10–36

GrawitchM,MunzD, Elliott E,Mathis A. 2003a. Promoting creativity in temporary problem-solving groups:the effects of positive mood and autonomy in problem definition on idea-generating performance.GroupDyn. 7:200–13

Grawitch MJ, Munz DC, Kramer TJ. 2003b. Effects of member mood states on creative performance intemporary workgroups. Group Dyn. 7:41–54

Hackman JR, Katz N. 2010. Group behavior and performance. In Handbook of Social Psychology, ed.ST Fiske, DT Gilbert, G Lindzey, pp. 1208–51. New York: Wiley. 5th ed.

Hareli S, Rafaeli A. 2008. Emotion cycles: on the social influence of emotion in organizations. Res. Organ.Behav. 28:35–59

Hatfield E, Cacioppo J, Rapson RL. 1993. Emotional contagion. Curr. Dir. Psychol. Sci. 2:96–99Hatfield E, Cacioppo J, Rapson RL. 1994. Emotional Contagion. Cambridge, UK: Cambridge Univ. PressHentschel T, Shemla M, Wegge J, Kearney E. 2013. Perceived diversity and team functioning: the role of

diversity beliefs and affect. Small Group Res. 44(1):33–61Hess U, Fischer A. 2014. Emotional mimicry: why and when we mimic emotions. Soc. Personal. Psychol.

Compass 8:45–57Hinsz VB, Tindale RS, Vollrath DA. 1997. The emerging conceptualization of groups as information pro-

cessors. Psychol. Bull. 121:43–64Hmieleski KM, Cole MS, Baron RA. 2012. Shared authentic leadership and new venture performance. J.

Manag. 38(5):1476–99

43www.annualreviews.org � Group Affect

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

Hochschild AR. 1983. The Managed Heart: Commercialization of Human Feelings. Berkeley: Univ. Calif.Press

Humphrey RH. 2002. The many faces of emotional leadership. Leadersh. Q. 13:493–504Ilies R, Judge TA. 2005. Goal regulation across time: the effects of feedback and affect. J. Appl. Psychol.

90:453–67Ilies R, Wagner DT, Morgeson FP. 2007. Explaining affective linkages in teams: individual differences in

susceptibility to contagion and individualism-collectivism. J. Appl. Psychol. 92:1140–48Isen AM. 2000. Positive affect and decision making. In Handbook of Emotions, ed. M Lewis, JM Haviland-

Jones, pp. 417–35. New York: Guilford. 2nd ed.Isen AM, Baron RA. 1991. Positive affect as a factor in organizational behavior.Res. Organ. Behav. 13:1–53JohnsonSK.2009.Doyou feelwhat I feel?Moodcontagionand leadershipoutcomes.Leadersh.Q.20:814–27Jones EE, Kelly JR. 2009. No pain, no gains: Negative mood leads to process gains in idea-generation groups.

Group Dyn. 13(2):75–88Kaplan S, LaPort K, Waller MJ. 2013. The role of positive affectivity in team effectiveness during crises. J.

Organ. Behav. 34:473–91Kelly JR, Barsade SG. 2001.Mood and emotions in small groups andwork teams.Organ. Behav.Hum.Decis.

Process. 86:99–130Keltner D, Haidt J. 1999. Social functions of emotions at four levels of analysis.Cogn. Emotion 13(5):505–21KlepAHM,WisseB, vander FlierH. 2011. Interactive affective sharing versus non-interactive affective sharing

inworkgroups: comparative effects of groupaffect onworkgroupperformance anddynamics.Eur. J. Soc.Psychol. 41(3):312–23

KlepAHM,Wisse B, van der FlierH. 2013.When sad groups expect tomeet again: interactive affective sharingand future interaction expectation as determinants of work groups’ analytical and creative task per-formance. Br. J. Soc. Psychol. 52(4):667–85. doi:10.1111/bjso.12000

Knight AP. 2015. Mood at the midpoint: affect and change in exploratory search over time in teams that facea deadline. Organ. Sci. 26:99–118

Knight AP, Baer M. 2014. Get up, stand up: the effects of a non-sedentary workspace on informationelaboration and group performance. Soc. Psychol. Personal. Sci. 5(8):910–17

Knight AP, Eisenkraft N. 2014. Positive is usually good, negative is not always bad: the effects of group affecton social integration and task performance. J. Appl. Psychol. In press

Kooij-de Bode HJM, van Knippenberg D, Van Ginkel WP. 2010. Good effects of bad feelings: negative af-fectivity and group decision-making. Br. J. Manag. 21:375–92

Kozlowski SWJ, Ilgen DR. 2006. Enhancing the effectiveness of work groups and teams. Psychol. Sci. PublicInterest 7:77–124

Lazarus RS. 1982. Thoughts on the relations between emotion and cognition. Am. Psychol. 37:1019–24Leidner R. 1993. Fast Food, Fast Talk: Service Work and the Routinization of Everyday Life. Berkeley: Univ.

Calif. PressLocke EA. 1976. The nature and causes of job satisfaction. In Handbook of Industrial and Organizational

Psychology, ed. MD Dunnette, pp. 1297–349. Chicago: Rand McNally. 1st ed.Lundqvist LO, Dimberg U. 1995. Facial expressions are contagious. J. Psychophysiol. 9:203–11Lyubomirsky S,KingL,DienerE. 2005.Thebenefits of frequent positive affect:Does happiness lead to success?

Psychol. Bull. 131(6):803–55Mackie DM, Devos T, Smith ER. 2000. Intergroup emotions: explaining offensive action tendencies in an

intergroup context. J. Personal. Soc. Psychol. 79:602–16Magee JC, Tiedens LZ. 2006. Emotional ties that bind: the roles of valence and consistency of group emotion

in inferences of cohesiveness and common fate. Personal. Soc. Psychol. Bull. 32:1703–15Mason CM,GriffinMA. 2003. Group absenteeism and positive affective tone: a longitudinal study. J. Organ.

Behav. 24:667–87McIntyre CW, Watson D, Clark LA, Cross SA. 1991. The effect of induced social interaction on positive and

negative affect. Bull. Psychon. Soc. 29:67–70Milliken FJ, Martins LL. 1996. Searching for common threads: understanding the multiple effects of diversity

in organizational groups. Acad. Manag. Rev. 21:402–33

44 Barsade � Knight

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

MoodyEJ,McIntoshDN,MannLJ,WeisserKR. 2007.More thanmeremimicry?The influence of emotiononrapid reactions to faces. Emotion 7:447–57

Moreland R. 1987. The formation of small groups. In Group Processes, ed. CA Hendrick, pp. 80–110. Rev.Personal. Soc. Psychol. Vol. 8. Thousand Oaks, CA: Sage

Niedenthal PM, Brauer M. 2012. Social functionality of human emotion. Annu. Rev. Psychol. 63:259–85O’Neill OA, Rothbard N. 2014. Love is all you need: debunking assumptions about masculinity and work-

family conflict through a multi-method study of emotional culture. Work. Pap., George Mason Univ.,Fairfax, VA

Parkinson B, Fischer AH, Manstead ASR. 2005. Emotion in Social Relations: Cultural, Group, and In-terpersonal Processes. New York: Psychology

Pastor JC, Mayo M, Shamir B. 2007. Adding fuel to the fire: the impact of followers’ arousal on ratings ofcharisma. J. Appl. Psychol. 92:1584–96

Pescosolido AT. 2002. Emergent leaders as managers of group emotion. Leadersh. Q. 13:583–99Picard RW. 2000. Affective Computing. Boston, MA: MIT PressPicard RW, Vyzas E, Healey J. 2001. Toward machine emotional intelligence: analysis of affective physio-

logical state. IEEE Trans. Pattern Anal. Mach. Intell. 23:1175–91Poh M, Swenson NC, Picard RW. 2010. A wearable sensor for unobtrusive, long-term assessment of elec-

trodermal activity. IEEE Trans. Biomed. Eng. 57:1243–52Rafaeli A, Sutton RI. 1989. The expression of emotion in organizational life. Res. Organ. Behav. 2:1–42Rhee SY. 2006. Shared emotions and group effectiveness: the role of broadening-and-building interactions.

Acad. Manag. Proc. (Meet. Abstr. Suppl.) Summer 2006:B1–6. doi: 10.5465/AMBPP.2006.22898617Rothbard NP, Wilk SL. 2011. Waking up on the right or wrong side of the bed: start-of-day workday mood,

work events, employee affect, and performance. Acad. Manag. J. 54:959–80Sanchez-Burks J, Huy QN. 2009. Emotional aperture and strategic change: the accurate recognition of

collective emotions. Organ. Sci. 20:22–34Schneider B. 1987. The people make the place. Person. Psychol. 40:437–53Seong JY, Choi JN. 2014. Effects of group-level fit on group conflict and performance: the initiating role of

leader positive affect. Group Organ. Manag. 39:190–212Shteynberg G, Hirsh JB, Apfelbaum EP, Larsen JT, Galinsky AD, Roese NJ. 2014a. Feeling more together:

Group attention intensifies emotion. Emotion 14:1102–14Shteynberg G, Hirsh JB, Galinsky AD, Knight AP. 2014b. Shared attention increases mood infusion. J. Exp.

Psychol. Gen. 143(1):123–30. doi:10.1037/a0031549Smith AC, Kleinman SK. 1989. Managing emotions in medical school: students’ contacts with the living and

dead. Soc. Psychol. Q. 52:56–69Sutton RI 1991. Maintaining norms about expressed emotions: the case of bill collectors. Adm. Sci. Q.

36:245–68Sy T, Choi JN. 2013. Contagious leaders and followers: exploring multi-stage mood contagion in a leader

activation andmember propagation (LAMP)model.Organ. Behav.Hum.Decis. Process. 122(2):127–40Sy T, Choi JN, Johnson SK. 2013. Reciprocal interactions between group perceptions of leader charisma and

group mood through mood contagion. Leadersh. Q. 24(4):463–76. doi:10.1016/j.leaqua.2013.02.002Sy T, Côté S, Saavedra R. 2005. The contagious leader: impact of the leader’s mood on the mood of group

members, group affective tone, and group processes. J. Appl. Psychol. 90:295–305Tanghe J,Wisse B, van der FlierH. 2010. The formation of group affect and team effectiveness: themoderating

role of identification. Br. J. Manag. 21:340–58Tiedens LZ, Sutton RI, Fong CT. 2004. Emotional variation in work groups: causes and performance

consequences. In The Social Life of Emotions, ed. LZ Tiedens, CW Leach, pp. 164–86. Cambridge, UK:Cambridge Univ. Press

Totterdell P. 2000.Catchingmoods and hitting runs:mood linkage and subjective performance in professionalsport teams. J. Appl. Psychol. 85:848–59

Totterdell P, Kellett S, Teuchmann K, Briner RB. 1998. Evidence of mood linkage in work groups. J. Personal.Soc. Psychol. 74:1504–15

45www.annualreviews.org � Group Affect

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

Totterdell P,Wall T,HolmanD,DiamondH, EpitropakiO. 2004. Affect networks: a structural analysis of therelationship between work ties and job-related affect. J. Appl. Psychol. 89:854–67

Tsai W, Chi N, Grandey AA, Fung A. 2012. Positive group affective tone and team creativity: negative groupaffective tone and team trust as boundary conditions. J. Organ. Behav. 33:638–56

Tuckman BW. 1965. Developmental sequence in small groups. Psychol. Bull. 63:384–99van der Schalk J, Fischer A, Doosje B,Wigboldus D, Hawk S, et al. 2011. Convergent and divergent responses

to emotional displays of ingroup and outgroup. Emotion 11:286–98Van Kleef GA. 2009. How emotions regulate social life. Curr. Dir. Psychol. Sci. 18(3):184–89Van Kleef GA, De Dreu CKW, Manstead ASR. 2010. An interpersonal approach to emotion in decision

making: the emotions as social information model. Adv. Exp. Soc. Psychol. 42:45–96Van Kleef GA, Homan AC, Beersma B, van Knippenberg D, van Knippenberg B, Damen F. 2009. Searing

sentiment or cold calculation? The effects of leader emotional displays on team performance depend onfollower epistemic motivation. Acad. Manag. J. 52:562–80

van Knippenberg D, Kooij-de Bode HJM, van Ginkel WP. 2010. The interactive effects of mood and traitnegative affect in group decision making. Organ. Sci. 21:731–44

van Knippenberg D, Schippers MC. 2007. Work group diversity. Annu. Rev. Psychol. 58:515–41Van Maanen J, Kunda G. 1989. Real feelings: emotional expression and organizational culture. Res. Organ.

Behav. 11:43–103Walter F, Bruch H. 2008. The positive group affect spiral: a dynamic model of the emergence of positive

affective similarity in work groups. J. Organ. Behav. 29:239–61Watson D. 2000. Mood and Temperament. New York: GuilfordWatsonD,Clark LA,McIntyreCW,Hamaker S. 1992.Affect, personality, and social activity. J. Personal. Soc.

Psychol. 63:1011–25WatsonD, Pennebaker JW. 1989.Health complaints, stress, and distress: exploring the central role of negative

affectivity. Psychol. Rev. 96:234–54Weisbuch M, Ambady N. 2008. Affective divergence: automatic responses to others’ emotions depend on

group membership. J. Personal. Soc. Psychol. 95:1063–79Weiss HM, Cropanzano R. 1996. Affective events theory: a theoretical discussion of the structure, causes and

consequences of affective experiences at work. Res. Organ. Behav. 18:1–74WilliamsKY,O’Reilly CA. 1998.Demography and diversity in organizations: a reviewof 40 years of research.

Res. Organ. Behav. 20:77–140YoungPT. 1961.Motivation andEmotion:ASurveyof theDeterminants ofHumanandAnimalActivity. New

York: WileyZajonc RB. 1980. Feeling and thinking: Preferences need no inferences. Am. Psychol. 35:151–75

46 Barsade � Knight

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

Annual Review of

Organizational

Psychology and

Organizational Behavior

Volume 2, 2015 Contents

Organizational Psychology Then and Now: Some ObservationsEdgar H. Schein . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Group AffectSigal G. Barsade and Andrew P. Knight . . . . . . . . . . . . . . . . . . . . . . . . . . 21

The Modeling and Assessment of Work PerformanceJohn P. Campbell and Brenton M. Wiernik . . . . . . . . . . . . . . . . . . . . . . . . 47

Justice, Fairness, and Employee ReactionsJason A. Colquitt and Kate P. Zipay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

Methodological and Substantive Issues in Conducting Multinational andCross-Cultural ResearchPaul E. Spector, Cong Liu, and Juan I. Sanchez . . . . . . . . . . . . . . . . . . . . 101

Leadership Development: An Outcome-Oriented Review Based on Time andLevels of AnalysesDavid V. Day and Lisa Dragoni . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

Beyond Lewin: Toward a Temporal Approximation of OrganizationDevelopment and ChangeJean M. Bartunek and Richard W. Woodman . . . . . . . . . . . . . . . . . . . . . 157

Beyond the Big Five: New Directions for Personality Research and Practice inOrganizationsLeaetta M. Hough, Frederick L. Oswald, and Jisoo Ock . . . . . . . . . . . . . 183

Corporate Social Responsibility: Psychological, Person-Centric, andProgressingDeborah E. Rupp and Drew B. Mallory . . . . . . . . . . . . . . . . . . . . . . . . . 211

Time in Individual-Level Organizational Studies: What Is It, How Is It Used,and Why Isn’t It Exploited More Often?Abbie J. Shipp and Michael S. Cole . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237

vi

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

Dynamics of Well-BeingSabine Sonnentag . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261

Low-Fidelity SimulationsJeff A. Weekley, Ben Hawkes, Nigel Guenole, and Robert E. Ployhart . . . 295

Emotional Labor at a Crossroads: Where Do We Go from Here?Alicia A. Grandey and Allison S. Gabriel . . . . . . . . . . . . . . . . . . . . . . . . . 323

Supporting the Aging Workforce: A Review and Recommendations forWorkplace Intervention ResearchDonald M. Truxillo, David M. Cadiz, and Leslie B. Hammer . . . . . . . . . 351

ESM 2.0: State of the Art and Future Potential of Experience SamplingMethods in Organizational ResearchDaniel J. Beal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383

Ethical LeadershipDeanne N. Den Hartog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409

Differential Validity and Differential Prediction of Cognitive Ability Tests:Understanding Test Bias in the Employment ContextChristopher M. Berry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435

Organizational Routines as Patterns of Action: Implications for OrganizationalBehaviorBrian T. Pentland and Thorvald Hærem . . . . . . . . . . . . . . . . . . . . . . . . . 465

Pay, Intrinsic Motivation, Extrinsic Motivation, Performance, and Creativityin the Workplace: Revisiting Long-Held BeliefsBarry Gerhart and Meiyu Fang . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489

Stereotype Threat in Organizations: Implications for Equity and PerformanceGregory M. Walton, Mary C. Murphy, and Ann Marie Ryan . . . . . . . . . . 523

Technology and Assessment in SelectionNancy T. Tippins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551

Workplace Stress Management Interventions and Health PromotionLois E. Tetrick and Carolyn J. Winslow . . . . . . . . . . . . . . . . . . . . . . . . . 583

Errata

An online log of corrections to Annual Review of Organizational Psychology andOrganizational Behavior articles may be found at http://www.annualreviews.org/errata/orgpsych.

Contents vii

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.

Ann

u. R

ev. O

rgan

. Psy

chol

. Org

an. B

ehav

. 201

5.2:

21-4

6. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

Acc

ess

prov

ided

by

50.5

8.14

7.12

0 on

04/

14/1

5. F

or p

erso

nal u

se o

nly.